Fuzzer test case results in 3 frames with dav1d vs 0 frames with libaom.

[0205/152809.960258:ERROR:dav1d_video_decoder.cc(205)] dav1d_send_data, ts: 0 s
[0205/152810.068615:ERROR:dav1d_video_decoder.cc(282)] dav1d_get_picture: 0s
[0205/152810.070064:ERROR:dav1d_video_decoder.cc(205)] dav1d_send_data, ts: 3e-05 s
[0205/152810.077857:ERROR:dav1d_video_decoder.cc(282)] dav1d_get_picture: 30us
<draining for EOS>
[0205/152810.151827:ERROR:dav1d_video_decoder.cc(282)] dav1d_get_picture: 30us
[0205/152810.151979:ERROR:dav1d_video_decoder.cc(98)] Error parsing OBU data

[0205/152809.909558:ERROR:aom_video_decoder.cc(205)] aom_codec_decode, ts: 0.388361 s
[0205/152809.926603:ERROR:aom_video_decoder.cc(205)] aom_codec_decode, ts: 0 s
<aom_codec_get_frame never returns frames for this file>

clusterfuzz-testcase-minimized-mediasource_MP4_AV1_pipeline_integration_fuzzer-4853788330557440

The code for dav1d_video_decoder can be seen here: https://cs.chromium.org/chromium/src/media/filters/dav1d_video_decoder.cc?&l=219

This isn't a huge deal AFAICT, so feel free to close as WontFix, but it seemed worth reporting. I haven't been able to get this to repro after converting the .mp4 to an ivf or using ffmpeg (which fails at some other stage).